Two things make most roadmaps useless. They carry dates that turn out to be wishes, and they list everything anybody ever asked for, which tells a reader nothing about what is actually happening.
So: no dates anywhere on this page, and a short list.
Now
The rebuild shipped at the end of August: one API instead of five services, Postgres instead of Mongo, and every outbound message passing through a single policy gate rather than three code paths that each made their own decisions. What is in hand now is the Android app, in testing with people who already use the product, and a weekly email that shows an account the questions it missed and the rule that would have answered them.
The parts that are done are the parts you use: the automation builder and its ten checks, the inbox with takeover, leads and CSV export, credits, voice profiles built from your own replies, knowledge sources, the admin controls, and export and erasure for your data. The security page describes how the last of those actually works.
Next
Three things, in roughly this order, and "roughly" is doing real work in that sentence.
- Mobile. Push notifications, the share sheet, and the product working properly on a phone — which is where most people read a comment in the first place.
- The growth loops the old build had and lost. Referrals, the waitlist, the newsletter. Unglamorous, and the reason people found us at all.
- Operability. Retrying a dead job without a database console, a metrics endpoint, and model routing that can be changed without a deploy. Nobody buys this. Everybody notices when it is missing.
Waiting on somebody else
WhatsApp. It is the most-asked-for thing by a distance, and the honest position is that we do not control the timeline. WhatsApp has its own approval process, its own message categories, and per-conversation charges from Meta that exist whatever a vendor charges on top. There is a waitlist inside the product; joining it is the only thing that gets you told first.
Not on the list, and not coming
This is the part worth reading, because a roadmap that only says yes is a sales document.
Decided against
Why
Three of those had screens in the old product, wired to endpoints that returned nothing real. Deleting them was the first decision of the rebuild, and it is the one we are most confident about.
What "next" is worth as a promise
Very little, and it is worth being explicit about why.
Everything in the Next section is genuinely started or genuinely queued, which is a real claim. It is not a claim that any of it lands in a particular month, or in that order, or at all. Work gets reordered by things nobody can schedule around — a platform change, a bug that turns out to be structural, a piece of compliance that arrives with a deadline attached.
The old product published a roadmap with quarters on it. Every quarter was wrong, and the cost was not embarrassment: it was people who chose us because something was listed for March, and then had to explain to somebody else why it had not arrived. That is the failure this page is shaped to avoid.
A date on a roadmap is a promise made by someone who does not yet know what they will find.
How this list changes
Not by a vote, and not by whoever asks loudest. What moves work up is the same question arriving repeatedly from people who have already connected an account and hit the thing they are asking about. That is a much smaller and much more useful signal than a feature request form.
If something here is wrong for you, say so at support@postengage.ai — the address goes to the people writing the code, and the about page explains why that is still true at this size.
When something ships it appears on the changelog, with the date it landed rather than the date it was hoped for.